Tomato Maturity Detection and Quality Grading Dataset

(1) Everyone wants to purchase high-quality, fresh fruits. People nowadays are so concerned about their health and conscious about what they should consume and what they should not. According to them, rotten fruits are detrimental to their health. As a result, the sale of fruits suffers which brings significant economic ramifications. One of the main reasons behind fruits getting defective is that the maturity detection process is still manual in Bangladesh and fruits would go rotten with the passage of time if it is not harvested at the proper time. To understand the proper harvesting time it is crucial to detect the mature and immature fruit properly. Tomato is one of the significant, widely consumed, well-liked, and nutrient-dense crops grown throughout Bangladesh. According to the estimation, in Bangladesh, every day there are significant financial losses because tomatoes become rotten easily. Hence, automated classification of mature, immature, fresh, and rotten tomato identification is indispensable to overcome this situation and bring out the benefit to fruit growers, retailers, and processing firms.

(2) In the recent era, computer vision techniques are very promising in performing such types of classification and detection tasks.

(3) With a view to developing computer vision-based algorithms, an extensive tomato dataset is presented containing Tomato Maturity Detection Dataset and Tomato Quality Grading Dataset. Tomato Maturity Detection Dataset consists of two classes namely immature and mature tomatoes whereas Tomato Quality Grading Dataset contains fresh and rotten tomatoes. The classifications of this dataset are done with the help of a domain expert from an agricultural institute.

(4) A total of 2986 images of mature, immature, fresh, and rotten tomatoes were collected from Sher-e-Bangla Agricultural University. Then from these original images, a total of 10,000 augmented images are produced by using rotation, zoom, flipping, and scaling techniques to increase the data number.